Reducing Hallucinations and Trade-Offs in Responses in Generative AI Chatbots for Cancer Information: Development and Evaluation Study
JMIR Cancer · 11.09.2025 · 11 · e70176-e70176
DOI: 10.2196/70176 ↗Kurz gefasst
Evaluationsstudie in der Zeitschrift JMIR Cancer, online erschienen am 11.09.2025. Verfasst von Sota Nishisako und 2 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Cancer Control, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo, 104-0045, Japan, 81 3 3547 5201, 81 3 3547 8577. Hauptthemen laut MEDLINE: Hallucinations, Neoplasms, Artificial Intelligence. Die Arbeit wurde bisher 38-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.
Abstract (Originalsprache)
Background
Generative artificial intelligence (AI) is increasingly used to find information. Providing accurate information is essential to support patients with cancer and their families; however, information returned by generative AIs is sometimes wrong. Returning wrong information is called hallucination. Retrieval-augmented generation (RAG), which supplements large language model (LLM) outputs with relevant external sources, has the potential to reduce hallucinations. Although RAG has been proposed as a promising technique, its real-world performance in public health communication remains underexplored.
Objective
This study aimed to examine cancer information returned by generative AIs with RAG using cancer-specific information sources and general internet searches to determine whether using RAG with reliable information sources reduces the hallucination rates of generative AI chatbots.
Methods
We developed 6 types of chatbots by combining 3 patterns of reference information with 2 versions of LLMs. Thus, GPT-4 and GPT-3.5 chatbots that use cancer information service (CIS) information, Google information, and no reference information (conventional chatbots) were developed. A total of 62 cancer-related questions in Japanese were compiled from public sources. All responses were generated automatically and independently reviewed by 2 experienced clinicians. The reviewers assessed the presence of hallucinations, defined as medically harmful or misinformation. We compared hallucination rates across chatbot types and calculated odds ratios (OR) using generalized linear mixed-effects models. Subgroup analyses were also performed based on whether questions were covered by CIS content.
Results
For the chatbots that used information from CIS, the hallucination rates were 0% for GPT-4 and 6% for GPT-3.5, whereas those for chatbots that used information from Google were 6% and 10% for GPT-4 and GPT-3.5, respectively. For questions on information that is not issued by CIS, the hallucination rates for Google-based chatbots were 19% for GPT-4 and 35% for GPT-3.5. The hallucination rates for conventional chatbots were approximately 40%. Using reference data from Google searches generated more hallucinations than using CIS data, with an OR of 9.4 (95% CI 1.2‐17.5, P
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute for Cancer Control, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo, 104-0045, Japan, 81 3 3547 5201, 81 3 3547 8577.
- Department of Public Health and Health Policy, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan.
- Headquarter for Cancer Information Services, National Cancer Center, Tokyo, Japan.
ORCID-Profile: Sota Nishisako, Takahiro Higashi, Fumihiko Wakao
Zitieren
APA
Nishisako, S., Higashi, T., & Wakao, F. (2025). Reducing Hallucinations and Trade-Offs in Responses in Generative AI Chatbots for Cancer Information: Development and Evaluation Study. JMIR Cancer, 11, e70176-e70176. https://doi.org/10.2196/70176
Vancouver
Nishisako S, Higashi T, Wakao F. Reducing Hallucinations and Trade-Offs in Responses in Generative AI Chatbots for Cancer Information: Development and Evaluation Study. JMIR Cancer. 2025;11:e70176-e70176. doi:10.2196/70176
BibTeX
@article{Nishisako2025,
author = {Nishisako, Sota and Higashi, Takahiro and Wakao, Fumihiko},
title = {Reducing Hallucinations and Trade-Offs in Responses in Generative AI Chatbots for Cancer Information: Development and Evaluation Study},
journal = {JMIR Cancer},
year = {2025},
volume = {11},
pages = {e70176--e70176},
doi = {10.2196/70176}
}
EUVIMED